Over-the-counter nicotine replacement therapy (OTC NRT) is, by far, the most common treatment for smoking cessation in the US. Over 80% of those using NRT will lapse and return to smoking. One possible reason for this high rate of treatment failure is that smokers are instructed to stop medication when they lapse (in contrast to recommendations to increase medication dose when heroin users lapse). Recent indirect evidence suggests that, in fact, a lapse is the most important time to continue NRT. We propose a randomized controlled trial of continued NRT post-lapse vs. stopping NRT post- lapse. Smokers who want to quit will receive, counseling, stop abruptly and begin NRT. All psychosocial treatment will occur via phone and medications via mail. Instructions and rationales for continuing or stopping NRT during a lapse episode will be delivered via written material, Interactive Voice Response (IVR) phone messages and reinforced during phone counseling. We will enter 1000 smokers and anticipate observing 580 lapses while using NRT. We hypothesize our intervention to continue NRT use upon a lapse will increase 6 month point prevalent abstinence with an OR of 2.0. We will also examine compliance with instructions and possible behavioral mechanisms of efficacy (e.g. does continued NRT reduce nicotine reward from cigarettes) as well as the incidence of serious adverse events when participants are concurrently smoking and using NRT after a lapse. This study will be the first direct experimental test of whether continuing NRT after a relapse increases abstinence and is safe. Positive results would suggest the package instructions for NRT, treatment guidelines and training programs should change to explicitly encourage smokers to continue NRT treatment after a lapse.